An instrument is described which is intended to perform a dual purpose (elemental-structural) analysis consistent with the environmental conditions implied by lunar or planetary operation. The dififractometer section is based on a modified Seeman-Bohlin focusing principle in which a sharp-line focus target, a powdered sample, and a movable detector slit all lie on the focusing circle. The convolution of the projections on the focal circle, of a narrow receiving slit on the detector, the line focus target, combined with a high dispersion produce higher resolution and intensity than is common with Bragg focusing diflfractorneters with simitar instrumental parameters. The range of d-spacings covered is from 1 to 7 Å (chromium target). The chemical analysis section of the instrument utilizes the fluorescent X-rays produced in the specimen by the primary beam. A proportional counter and pulse-height anaiyzer accomplish detection and energy discrimination. Resolution is low, but the analysis can distinguish between elements in the range of atomic numbers 11 to 29. Data from a breadboard model is presented. The entire unit, although primarily intended to meet the requirements of space, performs equally well as a routine laboratory analyzer. The horizontal, stationary nature of the specimen holder suggests several specific applications.

A two-crystal spectrometer that makes full use of conventional diffractometet: or spectrometer motions has been constructed so that it can be easily attached to a commercial instrument. The requisite degree of parallelism of the two crystalrotation axes is maintained by keeping both parallel to a single ground surface. Special alignment aids have been constructed to facilitate the rapid alignment of the instrument, which can be operated manually in 1″ intervals or used for automatic scanning in steps of 0.01°. Some guide lines for comparing the relative performance of spectrometers in X-ray absorption spectroscopy are suggested.

A two-crystal spectrometer employing a Cole polarizer made from a germanium crystal as one of the crystals is described. Experimental tests show that placing the polarizer crystal in the second position markedly improves the degree of plane polarisation, in full agreement with the predictions of the dynamical theory of X-ray diffraction. Tests indicate that the twice-diffracted beam h better than 95% plane polarized. The angular resolving power of the instrument is comparable to that of conventional two-crystal spectrometers. The almost wholly planepolarized monochromatic X-radiation, moreover, permits certain tests to be carried out far more effectively than is possible with partly polarized X-ray beams.

An appropriate test for homogeneity of composition of standards and of unknown samples is an analysis of variance of the intensity ratios of the components within the grains opposed to the variance between the grains under conditions of constant beam current capacitor change. Replication of analyses is necessary to distinguish inherent instrumental variation from actual chemical inhomogeneities present in the samples. This replication permits an estimate of the imprecision caused by in homogeneities and the statistical variation of X-ray quanta. The total uncertainty of an analysis is the sum of systematic error—primarily reference standard uncertainty—and the imprecision. For multiphase mixtures of microcrystalline refractory oxide material, an uncertainty of ±2% of the amount of oxide component present in a given phase should be considered a good analysis. This value is based on experimental work in the system MgO–FeO–Fe2O3–SiO2.

A “chopped beam” system of analysis has been devised for a scanning electron microprobe equipped with a 400-channel analyzer. The system has been programed with circuitry which places a signal onto the X or Y scanning coils of the probe, so that the beam jumps back and forth between two analytical areas. The same signal is used to activate alternate halves or quadrants on the 400-channel analyzer in a synchronous manner. The analyzer accumulates X-ray intensity data in the appropriate halves and quadrants as the electron beam oscillates between sample and standard or, in the case of trace analysis, between the unknown and the pure major constituent for background correction. The probe may be left unattended while it is gathering information in this manner. The dwell time of the probe on a given analytical area is 6 sec “live” time.

Errors due to instrumental drift and sample contamination are nullified or minimized by this technique. Consequently, theoretical precision is closely approached for extended counting times. 100 ppm levels of aluminum, silicon, nickel, and iron in uranium have been determined to precisions as good as ± 10 ppm at the 95% confidence level.

For the quantitative analysis of a 65 Cu-30 Ni-S Fe alloy, a 96 Cu-3 Si-1 Mn alloy, and a 78 Cu-20 Zn-2 Al alloy, the Ziebold empirical method of correcting electron-microbeam-probe data was used. Four binary standards, of single-phase Cu-Ni, Ni-Fe, Cu-Mn, and Cu-Zn alloys, were cast and the a correction factor found for each element in each binary by Ziebold's relationship (1 – K)/K – α (1 – C)/C, where K – I/I0 found in the probe and C is the weight fraction found by wet chemistry. The ARL EMX probe was used at 30 kV with a 25-μ beam diameter to negate inhomogeneities. Experience with these binaries indicated that in the presence of secondary fluorescence, the experimental α values agreed poorly with theoretically calculated K values; however, where secondary fluorescence was negligible, agreement between the experimental and theoretical α values was good. The α values for Cu–Si, Cu–Al, Al-Zn, and Mn–Si alioys, were therefore calculated from the theoretical equations. The α values for Cu–Fe alloys were also calculated from theoretical considerations because single-phase binaries over the composition range of interest could not be made for this system. All these α values were used in Ziebold's ternary equations to correct probe data (again using a 25-μ beam) from specimens of Cu–Ni-Fe, Cu–Si–Mn, and Cu-Zn–Al. These results were compared to wet-chemistry analyses for the same specimens with quite good correlation between the two sets of data. Calibration curves for the binary systems Cu-Ni, Cu-Fe, Ni-Fe, Cu-Mn, Cu-Si, Mn-Si, Cu-Al, Cu-Zn, and Al-Zn were made and are reproduced.

Most modern instrumental methods of analysis depend on the use of known standards of composition for calibration. Newer analytical techniques, such as the solids mass spectrometer, laser probe and, especially, the electron-probe microanalyzer have reduced the amount of a sample which can be analyzed quantitatively to a range of about 0.1 to as small as 0.00005 μg. As a corollary to these microanalytical advances, homogeneity requirements have become severe to meet analytical standards. This paper describes a continuation of the National Bureau of Standards' effort to characterize more fully existing standards as to suitability for the new microanalytical techniques. An NBS cartridge brass sample in both the wrought (NBS-1102) and chill cast forms (NBS-C1102), as well as a low-alloy steel sample (NBS-463), have been investigated by means of electron-probe micreanalysis and optical metallography. Some 17 elements are contained in the brass, while 25 elements are found in the steel. Results for 10 elements in the steel and 6 elements in the brass are presented. In the steel, iron, nickel, copper, and silicon ate essentially distributed homogeneously at micron levels, while manganese, tantalum, niobium, zirconium, sulfur, and chromium are not. In the brass, copper and zinc are distributed homogeneously at micron levels while lead, sulfur, aluminum, and silicon are not. Electron-probe micreanalyzer results indicate that both NBS-1102 and NBS-C1102 brass are suitable for use as a calibration standard for electronprobe microanalysis as well as other microsnalyticat techniques, such as the solids mass spectrometer. The results for brass have been corroborated by a number of laboratories using the electron-probe analyzer.

The electron probe is shown to be an effective tool for the analysis of the series of ferrimagnetic oxides referred to as the hexagonal ferrites. This series of compounds) containing barium, Fe3+, and a divalent metal cation, is formed by an ordered stacking of basic structural units in varying ratios. The ideal, complex stoichiornewy of these polytype-like mixed-layer structures can be computed from X-ray unit cell data; the various structures and their predicted stoichiometry are reviewed. Results of electron probe analysis of zinc-bearing single-crystal hexagonal ferrites are compared with theoretical values, the various correction procedures applied to the probe data are presented, and the limitations of the method in the analysis of hexagonal ferrites are discussed. The information obtained from this study has laid the groundwork for the determination of chemistry in substituted members of the hexagonal ferrite group, and will contribute significantly to the interpretation of the magnetic properties exhibited by these compounds.

Intensities and wavelengths of the K-series lines of silicon were measured using primary excitation in the electron probe. Specifically, intensity and line Position data for the Si K spectral lines α1, α2, ß, α3, and α4 were recorded using a high-resolution continuously curved mica-crystal vacuum spectrometer. Spectra were obtained from silicon metal and a wide range of mineral samples, e.g., SiO2, Mg2Si2O6, CaSiO3, KAlSi2O6, CaMgSi2O6, KAlSi3O8, NaAlSi3O8, CaAl2Si2O8, Fe2SiO4, and (Na, K) (AI, Si)2O4. Only slight differences were found to exist among the mineral spectra, but these differed markedly from that obtained from silicon metal.

An electron microprobe analyzer shielded and modified to accept for analysis radioactive samples reading up to 100 R/hr is described. The following features will be illustrated and discussed: (1) biological shielding, (2) detector shielding, (3) sample transfer system, (4) remote sample loading and focusing system, (5) laboratory layout, and (6) experimental results.

X-ray spectrochemical analysis is a versatile technique that can be used to determine considerably more than just the elemental composition of a sample. X-ray lines and bands are, in many cases, influenced by the state of chemical combination of the element whose spectrum is being investigated. Significant effects due to changes in bonding are seen in the K and L spectra of the low atomic number elements which fall in the 20 to 90 Å region. Results using primary excitation, a stearate crystal, and a flow proportional counter are shown for the K spectra of boron, carbon, and nitrogen. Chlorine and sulfur L spectra from some simple metal compounds of these elements are also shown.

A detailed study has been made of the intensities of the L, M, and N X-ray spectra for wavelengths in the region from 2-85 Å. Some of the lines of these spectra involve energy levels which are gradually being depopulated with decrease in atomic number, and this leads to a progressive simplification of the spectra as the wavelength increases. Spectra have also been obtained from several compounds involving these elements in different valence states. Differences in relative intensities have been observed for these compounds, but because of a lack of accurate information about absorption, it is impossible as yet to give such results an adequate interpretation.

An investigation has been made of the relative Kβ intensities in different chemical states of the sulfur atom using the Kα lines, with appropriate corrections, to provide the intensity standards. Both inorganic and organic compounds were included in the study. The data for each compound appear to be reliable to about ± 0.5%, while the whole series of compounds shows a variation greater than 20% in the corrected Kβ/Kα ratios. Energies were also measured, particularly the Kα energies, and their shifts were studied relative to the Kβ, intensity shifts. The work was done with a plane, single-crystal, helium-path spectrometer with proportional counter and pulse-height analysis for detection. The results are indicative of the usefulness of the method both in clarifying an uncertain chemical state and in studying the electronic structure of the bonded atom.

From measurements of the shift in wavelength with chemical combination, it appears there is very little transfer of charge between the metal and nonmetal atoms for the Group-IV transition-metal carbides and diborides. For Group-VI transition-metal carbides and beyond, there appears to be a definite transfer of charge from the carbon to the metal atom. The bonding in the Group-IV transitionmetal carbides is largely co valent, and metallic for the Group-IV transition-metal diborides. A comparison of shapes of emission bands with crystal structure indicates that, in general, NaCl-type structures have symmetrical peaks with narrow bandwidths compared to hexagonal structure, which have asymmetrical peaks, broad bandwidths, and long tails to the bottom of the bands.

Existing theories of X-ray absorption-edge fine structure do not adequately explain details of spectra observed for solids. However, the possibility that X-ray absorption spectra might eventually be used as tools for the characterizatiori of new insulator materials has prompted the study of an extensive selection of simple oxides of known chemistry and structure. The complete K absorption fine structure has been measured for some forty simple oxides of six elements of the fourth period. The L absorption-edge spectra have been measured for metallic lead and several lead oxides. The several isostructurat and polymorphic sets included among these oxides, as well as the Magnéli phases for three of the elements, have made it possible to study the effects of valence, coordination number, electron configuration, and crystal structure. The applicability of current theories of the fine structures are discussed in the light of these findings. An automated single-crystal spectrometer is described.

A pragmatic, easily applied theory of the extended X-ray absorption fine structure found on the high-energy side of absorption edges is developed. This simple model can be successfully applied to determinations of interatomic distances, and it accounts for differences in Kronig structure on different absorption edges in the same material. Quantitative agreement with experimental data is good.

X-ray work in the 10 to 400 Å range usually requires, for maximum sensitivity, new methods of X-ray generation, analysis, and detection. A comparative survey includes X-ray generation by fluorescence, electron bombardment, and proton bombardment. Dispersive analysis with crystals and gratings is described. Finally, window less photoelectric multipliers are evaluated as X-ray detectors.

Results show that typical efficiences range from 10−4 to 10−2 photons/electron or photons/proton, with the latter value as a practical upper limit. An efficiency of 8.5% has been measured for a KAP crystal covering the wavelength range up to 25 Å, while an original gold-plated grating lias a maximum measured efficiency of 20%. Counter efficiency may range from 1 to 50%, depending on wavelength. Total efficiency for a grating-type spectrometer in the 20 to 400 Å range has been measured to 2 × 105 photons/cm2-count.

The importance of operating within the blaze angle is emphasized by a series of measurements of ultrasoft X-rays obtained with a new grazing-incidence, grating spectrometer, A formula to maximize output intensity has been derived which predicts a variation of input angle with changes of wavelength. The spectrometer is designed to utilize this function. The grating spectrometer was installed into an Applied Research Laboratories Analyst's Microprobe. Since the AMX furnishes an utt.rasoft spectrometer with a fatty acid pseudocrystal as standard equipment, it is possible to compare the two systems simultaneously on the same sample at the same wavelength. The grating and crystal systems are compared for peak intensity and line-to-background ratio.

Methods and instruments have been developed for efficient spectroscopic analysis in the 15 to 150 Å ultrasoft X-ray legion using specially designed Langmuir-Blodgett type multilayer analyzers with a flow proportional counter detection system. Simple adaptations of the basic Philips vacuum spectrograph are employed with a high-intensity, demountable, ultrasoft X-ray source to excite fluorescence in the lighter elements for chemical and valence band analysis.